tooluniverse-structural-variant-analysis

Pass

Audited by Gen Agent Trust Hub on Apr 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed for clinical genomics interpretation and follows a rigorous 7-phase workflow. All described activities are consistent with the stated purpose of analyzing structural variants.
  • [COMMAND_EXECUTION]: The skill includes Python code snippets in SKILL.md and REFERENCE.md. These snippets serve as templates for the agent to interact with a specialized tooluniverse library. The logic is restricted to querying scientific databases (e.g., ClinGen, OMIM, gnomAD) and performing data processing. No arbitrary shell commands or malicious execution patterns were detected.
  • [EXTERNAL_DOWNLOADS]: The skill references several well-known scientific databases and services, including NCBI, Ensembl, ClinVar, DECIPHER, and PubMed. These are legitimate, trusted sources for genomic data. No suspicious external script downloads or unverified package installations are present.
  • [DATA_EXFILTRATION]: There is no evidence of unauthorized data transmission or credential harvesting. The network operations are limited to standard API queries for clinical research data.
  • [PROMPT_INJECTION]: The instructions focus on structured analysis and reporting. No attempts to override agent safety protocols or bypass constraints were found.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 16, 2026, 08:27 AM